Within medicinal chemistry there continues to be a need for new synthetic transformations with high selectivity. Since enzymatic reactions typically exhibit good selectivities, we have been interested in developing biomimetic reactions which imitate the style of enzymatic transformations and, we hope, their effectiveness. 1. Understanding the types of templates that can bind and deliver free radicals to specific hydrogens. Our discovery that nitrogen heterocycles can be used opens up a new area. 2. Learning how to make C-O, C-F, and C-N bonds. They are of significant importance in medicinal chemistry. 3. The use of catalytic amounts of template molecules. By the use of ion pairing, hydrogen bonding, metal coordination, or molecular complexing we plan to produce temporary associations between catalysts and substrates. 4. The introduction of multiple interactions between substrate and reagent. Greatly increased selectivity should result, including enantioselectivity. 5. The selective functionalization of flexible molecules. Multiple interactions can immobilize an otherwise disordered substrate. 6. The functionalization of even less reactive C-H bonds. With appropriate orientation, we should be able to make geometry the only factor governing attack by a ferocious reagent. 7. Generalization of the types of transformations which can be geometrically directed. Particularly attractive is the development of metal-porphyrin catalyzed chemistry, including a nitrogen analog of the cytochrome P-450 model system which we have recently discovered. 8. The use of molecular modelling techniques to correlate and design template- directed reactions.